Pixelmath + Starnet Question

Discuss how you are able to get those fantastic images!!!
Post Reply
User avatar
Baskevo
Orion Spur Ambassador
Articles: 0
Offline
Posts: 829
Joined: Sun Aug 11, 2019 8:47 am
4
Location: Orange County, California
Status:
Offline

TSS Photo of the Day

Pixelmath + Starnet Question

#1

Post by Baskevo »


Hi guys!

I'm working on a new project, and half of my data is crap because of the stars... I'm working on a fix, but for now, I have an idea!

So, when I stacked all of my subs without cutting anything, I get quite a lot more detail in the object, but again the stars are terrible. If I take out half of the subs, I don't get a lot of signal, but my stars look great! So...

What if I use Starnet++ to remove the stars from the high-signal image, then combine it with the stars from the low-signal image? Does anyone have any suggestions on how I could do that?

thanks in advance!
Clear skies :)
-James W.

Telescope: Explore Scientific 80mm FCD100 Triplet APO Refractor
Mount: EQ6-R Pro
Cameras: ZWO ASI1600mm Pro (Cooled) | Canon DSLR EOS T7i
Auto-guiding: ZWO ASI120mm-Mini + Astromania 50mm Guidescope

Filters: ZWO 31mm Ha/Oiii/Sii 7nm + LRGB | Orion 2" Skyglow Filter
Accessories: Explore Scientific 2" Field Flattener, ZWO EFW 8 Position
Software: APT, SharpCap Pro, PHD2, CPWI | PixInsight, DeepSkyStacker, Photoshop

Flickr: https://www.flickr.com/gp/186194203@N06/18B629
User avatar
Stuart United States of America
Jupiter Ambassador
Articles: 0
Offline
Posts: 247
Joined: Sat Jun 01, 2019 11:12 pm
4
Location: East Bay, California
Status:
Offline

TSS Photo of the Day

Re: Pixelmath + Starnet Question

#2

Post by Stuart »


Give it a shot.
Personal equipment: TEC 140 F7 on Astro-Physics Mach 1 mount. Camera QSI 683ws7. Guide with Vario guiding scope
Shared equipment through Star Shadows Remote Observatory through PROMPT/ CTIO/Chile 16" RCOS 16803 chip
Shared equipment through San Diego Astronomy Society 14" RC with 16803 chip on a paramount
Software (for my stuff) PemPRO, SGP, PHD, Focus Boss, ASCOM, and Pixinsight on the other end.
User avatar
Baskevo
Orion Spur Ambassador
Articles: 0
Offline
Posts: 829
Joined: Sun Aug 11, 2019 8:47 am
4
Location: Orange County, California
Status:
Offline

TSS Photo of the Day

Re: Pixelmath + Starnet Question

#3

Post by Baskevo »


How would I do it, is my question :)
-James W.

Telescope: Explore Scientific 80mm FCD100 Triplet APO Refractor
Mount: EQ6-R Pro
Cameras: ZWO ASI1600mm Pro (Cooled) | Canon DSLR EOS T7i
Auto-guiding: ZWO ASI120mm-Mini + Astromania 50mm Guidescope

Filters: ZWO 31mm Ha/Oiii/Sii 7nm + LRGB | Orion 2" Skyglow Filter
Accessories: Explore Scientific 2" Field Flattener, ZWO EFW 8 Position
Software: APT, SharpCap Pro, PHD2, CPWI | PixInsight, DeepSkyStacker, Photoshop

Flickr: https://www.flickr.com/gp/186194203@N06/18B629
User avatar
Stuart United States of America
Jupiter Ambassador
Articles: 0
Offline
Posts: 247
Joined: Sat Jun 01, 2019 11:12 pm
4
Location: East Bay, California
Status:
Offline

TSS Photo of the Day

Re: Pixelmath + Starnet Question

#4

Post by Stuart »


Ah. Okay. Here is a general way to do it. If you want a step by step workflow, I can provide it but I'll have to do it later when I have more time. Basically what you need to do is process both images, and then through a star mask apply the stars to the starless image.

Let me know if you want more step-by-step instructions. I can provide them.
Personal equipment: TEC 140 F7 on Astro-Physics Mach 1 mount. Camera QSI 683ws7. Guide with Vario guiding scope
Shared equipment through Star Shadows Remote Observatory through PROMPT/ CTIO/Chile 16" RCOS 16803 chip
Shared equipment through San Diego Astronomy Society 14" RC with 16803 chip on a paramount
Software (for my stuff) PemPRO, SGP, PHD, Focus Boss, ASCOM, and Pixinsight on the other end.
User avatar
Baskevo
Orion Spur Ambassador
Articles: 0
Offline
Posts: 829
Joined: Sun Aug 11, 2019 8:47 am
4
Location: Orange County, California
Status:
Offline

TSS Photo of the Day

Re: Pixelmath + Starnet Question

#5

Post by Baskevo »


Stuart wrote: Thu Feb 20, 2020 7:00 pm Ah. Okay. Here is a general way to do it. If you want a step by step workflow, I can provide it but I'll have to do it later when I have more time. Basically what you need to do is process both images, and then through a star mask apply the stars to the starless image.

Let me know if you want more step-by-step instructions. I can provide them.
Thanks Stuart... Do I just take a straight star mask, and then blend it in like an HDR image with the high-signal image? I will give it a shot and see what I can do... Thanks for the help :)
-James W.

Telescope: Explore Scientific 80mm FCD100 Triplet APO Refractor
Mount: EQ6-R Pro
Cameras: ZWO ASI1600mm Pro (Cooled) | Canon DSLR EOS T7i
Auto-guiding: ZWO ASI120mm-Mini + Astromania 50mm Guidescope

Filters: ZWO 31mm Ha/Oiii/Sii 7nm + LRGB | Orion 2" Skyglow Filter
Accessories: Explore Scientific 2" Field Flattener, ZWO EFW 8 Position
Software: APT, SharpCap Pro, PHD2, CPWI | PixInsight, DeepSkyStacker, Photoshop

Flickr: https://www.flickr.com/gp/186194203@N06/18B629
User avatar
Stuart United States of America
Jupiter Ambassador
Articles: 0
Offline
Posts: 247
Joined: Sat Jun 01, 2019 11:12 pm
4
Location: East Bay, California
Status:
Offline

TSS Photo of the Day

Re: Pixelmath + Starnet Question

#6

Post by Stuart »


Hey James,
No. I'll put together a workflow for you in a bit. I just need to get to my Pixinsight machine. I'm outside right now rewiring my mount. :-)
Stu
Personal equipment: TEC 140 F7 on Astro-Physics Mach 1 mount. Camera QSI 683ws7. Guide with Vario guiding scope
Shared equipment through Star Shadows Remote Observatory through PROMPT/ CTIO/Chile 16" RCOS 16803 chip
Shared equipment through San Diego Astronomy Society 14" RC with 16803 chip on a paramount
Software (for my stuff) PemPRO, SGP, PHD, Focus Boss, ASCOM, and Pixinsight on the other end.
User avatar
KathyNS Canada
Co-Administrator
Co-Administrator
Articles: 0
Offline
Posts: 2584
Joined: Thu Apr 25, 2019 11:47 am
4
Location: Nova Scotia
Status:
Offline

TSS Awards Badges

TSS Photo of the Day

Re: Pixelmath + Starnet Question

#7

Post by KathyNS »


My technique is to apply Starnet to a clone of the image. So I have an 'Original' and a 'Starless' image. Then I use PixelMath to subtract 'Starless' from 'Original', to give 'StarsOnly'. I process the target ('Starless') as required, and I process the stars ('StarsOnly') as required. I can freely use Morphological Transfer on 'StarsOnly' because it will leave no artifacts on the background, since there isn't any. Finally, I use PixelMath again to add 'Starless' and 'StarsOnly' to create an 'Combined' image.
Image
DSO AP: Orion 200mm f/4 Newtonian Astrograph; ATIK 383L+; EFW2 filter wheel; Astrodon Ha,Oiii,LRGB filters; KWIQ/QHY5 guide scope; Planetary AP: Celestron C-11; ZWO ASI120MC; Portable: Celestron C-8 on HEQ5 pro; C-90 on wedge; 20x80 binos; Etc: Canon 350D; Various EPs, etc. Obs: 8' Exploradome; iOptron CEM60 (pier); Helena Observatory (H2O) Astrobin
User avatar
Baskevo
Orion Spur Ambassador
Articles: 0
Offline
Posts: 829
Joined: Sun Aug 11, 2019 8:47 am
4
Location: Orange County, California
Status:
Offline

TSS Photo of the Day

Re: Pixelmath + Starnet Question

#8

Post by Baskevo »


KathyNS wrote: Thu Feb 20, 2020 11:52 pm My technique is to apply Starnet to a clone of the image. So I have an 'Original' and a 'Starless' image. Then I use PixelMath to subtract 'Starless' from 'Original', to give 'StarsOnly'. I process the target ('Starless') as required, and I process the stars ('StarsOnly') as required. I can freely use Morphological Transfer on 'StarsOnly' because it will leave no artifacts on the background, since there isn't any. Finally, I use PixelMath again to add 'Starless' and 'StarsOnly' to create an 'Combined' image.
Beautiful!! Thank you for this, this is great! :) Can't wait to try it! Stay tuned :)
-James W.

Telescope: Explore Scientific 80mm FCD100 Triplet APO Refractor
Mount: EQ6-R Pro
Cameras: ZWO ASI1600mm Pro (Cooled) | Canon DSLR EOS T7i
Auto-guiding: ZWO ASI120mm-Mini + Astromania 50mm Guidescope

Filters: ZWO 31mm Ha/Oiii/Sii 7nm + LRGB | Orion 2" Skyglow Filter
Accessories: Explore Scientific 2" Field Flattener, ZWO EFW 8 Position
Software: APT, SharpCap Pro, PHD2, CPWI | PixInsight, DeepSkyStacker, Photoshop

Flickr: https://www.flickr.com/gp/186194203@N06/18B629
User avatar
Stuart United States of America
Jupiter Ambassador
Articles: 0
Offline
Posts: 247
Joined: Sat Jun 01, 2019 11:12 pm
4
Location: East Bay, California
Status:
Offline

TSS Photo of the Day

Re: Pixelmath + Starnet Question

#9

Post by Stuart »


Okay, sorry for the delay. If Kathy's way works for you, great. I recommend using the max operator as opposed to addition, but that's just me.

The way I do it is a little more convoluted, and it involves making a perfect star mask.
1) Take your STARS image and make a star mask of it. Make sure the star mask has ALL the stars that you are interested in. This may mean doing different star masks at different levels, and then using the max operator to put them all together for example, max(starmask1, starmask2)

2) On your starnet image, apply the star mask you built

3) On your star image, do channel extraction. You can do Lab or RGB.

4) On your starnet image with the starmask applied, do channel combination with either RGB or Lab (depending on what you did to extract) and assign the extracted images that you did on step 3.

Hope that helps
Stu
Personal equipment: TEC 140 F7 on Astro-Physics Mach 1 mount. Camera QSI 683ws7. Guide with Vario guiding scope
Shared equipment through Star Shadows Remote Observatory through PROMPT/ CTIO/Chile 16" RCOS 16803 chip
Shared equipment through San Diego Astronomy Society 14" RC with 16803 chip on a paramount
Software (for my stuff) PemPRO, SGP, PHD, Focus Boss, ASCOM, and Pixinsight on the other end.
User avatar
KathyNS Canada
Co-Administrator
Co-Administrator
Articles: 0
Offline
Posts: 2584
Joined: Thu Apr 25, 2019 11:47 am
4
Location: Nova Scotia
Status:
Offline

TSS Awards Badges

TSS Photo of the Day

Re: Pixelmath + Starnet Question

#10

Post by KathyNS »


Stuart wrote: Fri Feb 21, 2020 6:07 amI recommend using the max operator as opposed to addition, but that's just me.
If it works for you, great. I'd be cautious about recombining the images with Max if the image is heavy with nebula. When you do the subtraction to get the stars-only image, you will be subtracting some nebula colour (typically red) from the stars, causing a colour shift in stars that were in front of the nebula. This is a limitation of my method. It is not a problem if you recombine with addition, since the nebula colour gets added back in. The Max operation would not restore the star colours.

If you use Max to recombine, you might want to come up with a better strategy for separating the stars that doesn't give a colour shift.
Image
DSO AP: Orion 200mm f/4 Newtonian Astrograph; ATIK 383L+; EFW2 filter wheel; Astrodon Ha,Oiii,LRGB filters; KWIQ/QHY5 guide scope; Planetary AP: Celestron C-11; ZWO ASI120MC; Portable: Celestron C-8 on HEQ5 pro; C-90 on wedge; 20x80 binos; Etc: Canon 350D; Various EPs, etc. Obs: 8' Exploradome; iOptron CEM60 (pier); Helena Observatory (H2O) Astrobin
User avatar
Stuart United States of America
Jupiter Ambassador
Articles: 0
Offline
Posts: 247
Joined: Sat Jun 01, 2019 11:12 pm
4
Location: East Bay, California
Status:
Offline

TSS Photo of the Day

Re: Pixelmath + Starnet Question

#11

Post by Stuart »


KathyNS wrote: Fri Feb 21, 2020 12:47 pm
Stuart wrote: Fri Feb 21, 2020 6:07 amI recommend using the max operator as opposed to addition, but that's just me.
If it works for you, great. I'd be cautious about recombining the images with Max if the image is heavy with nebula. When you do the subtraction to get the stars-only image, you will be subtracting some nebula colour (typically red) from the stars, causing a colour shift in stars that were in front of the nebula. This is a limitation of my method. It is not a problem if you recombine with addition, since the nebula colour gets added back in. The Max operation would not restore the star colours.

If you use Max to recombine, you might want to come up with a better strategy for separating the stars that doesn't give a colour shift.
Completely agree. The danger of combining by simple addition is that you might get a value greater than 1. If you rescale then it should be fine.
To do it my way, you have to have a REALLY careful star mask.
Personal equipment: TEC 140 F7 on Astro-Physics Mach 1 mount. Camera QSI 683ws7. Guide with Vario guiding scope
Shared equipment through Star Shadows Remote Observatory through PROMPT/ CTIO/Chile 16" RCOS 16803 chip
Shared equipment through San Diego Astronomy Society 14" RC with 16803 chip on a paramount
Software (for my stuff) PemPRO, SGP, PHD, Focus Boss, ASCOM, and Pixinsight on the other end.
User avatar
STEVE333 United States of America
Inter-Galactic Ambassador
Articles: 0
Offline
Posts: 3466
Joined: Sat May 11, 2019 5:01 pm
4
Location: Santa Cruz, Ca, USA
Status:
Offline

TSS Awards Badges

TSS Photo of the Day

Re: Pixelmath + Starnet Question

#12

Post by STEVE333 »


Hi James - Here is a third way to consider (just in case you aren't thoroughly confused :confusion-confused: ).
1) Produce your version with good stars (call this one Stars), but, don't push the nebulosity too hard (don't stretch too far).
2) Produce your version with good nebulosity. While this version is still "mildly" stretched, produce the starless version from this (call it Starless).
3) If the starless version has blotches where bright stars were I will often gently CloneStamp out the blotches. Continue stretching, sharpening, etc. the starless version.
4) Put the stars back into the processed starless version using PixelMath expressions as shown below. You can adjust "f" to taste. I typically use a value of "f" in the range 0.9 - 1. Too low a value of "f" will dim the stars. Just click on the square (Apply) button to produce the result.

Image

That's it.

Cheers,
Steve
Steve King: Light Pollution (Bortle 5)
Telescope + Mount + Guiding: W.O. Star71-ii + iOptron CEM40 EC + Orion Magnificent Mini AutoGuider
Camera: ASI 1600MM Pro + EFW Filter Wheel + Chroma 3nm Siii, Ha, Oiii + ZWO LRGB Filters
Software: PHD2; APT; PixInsight ***** My AP website: www.steveking.pictures
Image
Image
User avatar
Baskevo
Orion Spur Ambassador
Articles: 0
Offline
Posts: 829
Joined: Sun Aug 11, 2019 8:47 am
4
Location: Orange County, California
Status:
Offline

TSS Photo of the Day

Re: Pixelmath + Starnet Question

#13

Post by Baskevo »


STEVE333 wrote: Mon Feb 24, 2020 4:54 pm Hi James - Here is a third way to consider (just in case you aren't thoroughly confused :confusion-confused: ).
1) Produce your version with good stars (call this one Stars), but, don't push the nebulosity too hard (don't stretch too far).
2) Produce your version with good nebulosity. While this version is still "mildly" stretched, produce the starless version from this (call it Starless).
3) If the starless version has blotches where bright stars were I will often gently CloneStamp out the blotches. Continue stretching, sharpening, etc. the starless version.
4) Put the stars back into the processed starless version using PixelMath expressions as shown below. You can adjust "f" to taste. I typically use a value of "f" in the range 0.9 - 1. Too low a value of "f" will dim the stars. Just click on the square (Apply) button to produce the result.

Image

That's it.

Cheers,
Steve


Sweet thank you Steve! This is essentially what I tried, which I got from your original starlet posts...It did not work too well though, because the good stars were much smaller, and it looked fake... I could try the clone stamping, but that sounds like too much work to be honest! I will give it another shot on my next image...

Thanks for the ideas, everyone! It is probably easier for me to just capture better subs next time :)
-James W.

Telescope: Explore Scientific 80mm FCD100 Triplet APO Refractor
Mount: EQ6-R Pro
Cameras: ZWO ASI1600mm Pro (Cooled) | Canon DSLR EOS T7i
Auto-guiding: ZWO ASI120mm-Mini + Astromania 50mm Guidescope

Filters: ZWO 31mm Ha/Oiii/Sii 7nm + LRGB | Orion 2" Skyglow Filter
Accessories: Explore Scientific 2" Field Flattener, ZWO EFW 8 Position
Software: APT, SharpCap Pro, PHD2, CPWI | PixInsight, DeepSkyStacker, Photoshop

Flickr: https://www.flickr.com/gp/186194203@N06/18B629
Post Reply

Create an account or sign in to join the discussion

You need to be a member in order to post a reply

Create an account

Not a member? register to join our community
Members can start their own topics & subscribe to topics
It’s free and only takes a minute

Register

Sign in

Return to “Image processing”